[ 1 65 ] 
He filled fix glaffes with fnow, as ufual, and put the 
thermometer in one of them, pouring thereupon the 
fpirit of nitre. When the mercury would fall no 
lower in this, he, in the fame manner, put it in a 
fecond, then in a third, and fo in a fourth j in which 
fourth immerfion, the mercury was congealed. 
Another very confiderable difference prefented it- 
felf in purfuing thefe inquiries, with regard to the 
mode of defcent of the mercury. He conftantly and 
invariably obferved, that the mercury defcended at 
firft gently, but afterwards very rapidly. But the 
point, at which this impetus begins, is not eafy to 
afcertain ; as in different experiments it begins very 
differently, and fometimes at about 300, at other 
times about 350, and even further. In the experi- 
ment before-mentioned, in which the mercury fell 
to 800, it proceeded very regularly to 600 ; about 
which point it began to defcend, with very great 
fwiftnefs, and the bulb of the thermometer was 
broke. The mercury, however, was perfectly con- 
gealed. 
He frequently obferved another remarkable phe- 
nomenon; which was, that although the fpirit of 
nitre, the fnow, and the mercury in the thermome- 
ter, were previoufly reduced to the fame temperature, 
upon pouring the fpirit of nitre upon the fnow, the 
mercury in the thermometer rofe. But as this did not 
always happen, he carefully attended to every cir- 
cumftance; from which it appeared, that this effect 
arofe from his pouring the aquafortis immediately 
upon the bulb of the thermometer, not previoufly 
well immerfed in the fnow. He likewife obferved 
another effect, twice only ; and this was, that, after 
the 
